Download Planning for PACS - American College of Radiology

Planning for PACS: A Comprehensive
Guide to Nontechnical Considerations
Mervyn D. Cohen, MB, ChB, MD, Lori L. Rumreich, MBA, Kimberley M. Garriot,
S. Gregory Jennings, MD
A complete picture archiving and communication system (PACS) installation is one of the largest projects a
radiology department will undertake. Although technology issues are important, they often draw focus away
from many other significant issues This paper describes in detail all of these other necessary components that
need to be addressed if a PACS installation is to be relatively trouble free, provides guidelines for successful
PACS implementation, and details pitfalls to be avoided.
Key Words: Process reengineering, PACS, radiology economics, PACS installation
J Am Coll Radiol 2005;2:327-337. Copyright © 2005 American College of Radiology
INTRODUCTION
When considering a picture archiving and communication system (PACS) implementation, one first seeks a
viable economic justification. If this is found, one next
identifies the best hardware and software and develops a
robust network that will drive the PACS system. All are
important elements of a PACS implementation, but a
successful PACS program requires careful and deliberate
attention to many other critical issues.
This review does not deal with system measures (e.g.,
compression, transmission rates, storage capacity) or the
detailed technical specifications that must be met to ensure a smoothly functioning PACS; several good articles
exist [1–3]. These technical issues may overshadow other
equally important considerations. This article thus is not
a comprehensive review of the PACS literature but recounts personal experiences and focuses on many of the
often neglected smaller items that are essential for a satisfactory PACS purchase and installation.
FINANCIAL JUSTIFICATION
The idea to implement a PACS in an organization often
comes from a radiologist. The typical institutional response is moderate interest, followed by demands to justify the significant cost. A PACS has obvious costs (hardware, software, and service contracts) but also obvious
savings (a reduction in personnel and the elimination of
Department of Radiology, Riley Hospital for Children, Indiana University
School of Medicine, Indianapolis, Indiana.
Corresponding author and reprints: Mervyn D. Cohen, MB, ChB, MD,
Department of Radiology, Riley Hospital for Children, 702 Barnhill Dr.,
Room 1053, Indianapolis, IN 46202; e-mail: [email protected].
© 2005 American College of Radiology
0091-2182/05/$30.00 ● DOI 10.1016/j.jacr.2004.08.024
film and film processing). Unfortunately, neither the
costs nor the savings can be presented in such a simplistic
fashion.
Capital costs include hardware and software, network
equipment, Digital Imaging and Communications in
Medicine (DICOM) upgrades, computed radiography
systems, and facility modifications [4]. Every institution
will begin PACS evaluation at a different point in its
evolution from analog to digital imaging. Some may
already acquire nearly all radiographic images digitally,
with extensive computed radiography systems and digital
fluoroscopic units, and have DICOM-compliant magnetic resonance (MR) and computed tomography (CT)
systems. But if an organization lacks computed radiography and has MR and CT systems that are not DICOM
compliant, it will have to spend heavily to acquire these
systems, and these costs will be added to its PACS program. The presence of preexisting computed radiography
systems or direct digital systems for plain-film radiography can determine whether PACS installation is potentially profitable [5]. Other items may be included in the
costs of a PACS program. For example, some institutions
may already have sophisticated networking capabilities
that require little expansion to be able to support the
demands of a PACS, but others will need to develop these
systems. Some institutions may have extensive information technology (IT) resources available, whereas others
may have to expand these resources. Cost assigned to the
PACS project will also vary depending on whether other
costs are included (e.g., hardware and software upgrades,
service maintenance contracts, system obsolescence).
Clearly, it is quite difficult, if not impossible, to identify the true costs of PACS implementation, and such
costs are very site specific. Financial analysis of the PACS
327
328 Journal of the American College of Radiology/ Vol. 2 No. 4 April 2005
Table 1. Financial benefits of a picture archiving
and communication system
Real, tangible dollar ● Reduction of personnel
savings
● Elimination of film and film
processing
● Film loan space available
for other use
Time value of
● Improved technologist
money
productivity
● Improved radiologist
productivity
● Improved referring
physician productivity
Qualitative value of ● Improved report
money
turnaround time
● Decreased length of
patient stay in the hospital
● Improved patient
satisfaction
● Improved referring
physician satisfaction
benefit is further complicated by a trend of continually
declining purchase costs, causing analyses to quickly become out of date [4].
The savings generated from implementing a PACS are
even more difficult to quantify. Financial benefits from a
PACS may be placed in three broad categories (see Table
1). Savings should include reduced film library space,
warehouse cost, the cost of film and development supplies, salary reductions from personnel cuts, and the potential replacement of equipment such as film processors
[4]. However, even these tangible dollar savings may not
be realized. For example, personnel may not be decreased, and old film storage space may not be fully used.
Studies have clearly demonstrated improved technologist productivity after a PACS installation; the time
required to perform ultrasound studies decreased [6],
and the time for performing CT studies was decreased by
as much as 45% [7]. However, the increase in technologist productivity may be gained in small segments of time
that are not fully used to increase patient throughput or
to decrease staffing, and the potential cost savings may
thus not be realized [8].
The qualitative savings may be extremely important,
but it can be difficult to assign measurable dollar
amounts to them. Some of the intangible benefits of a
PACS may indeed be of more value to an institution than
some of the tangible dollar savings. A PACS clearly delivers benefits to multiple departments and improves the
overall quality of patient care [4].
It is thus very apparent that a decision to move ahead
with the implementation of a PACS program cannot and
should not be based purely on the analysis of cost and
savings. These can be presented to create almost any
predetermined economic result and may also miss many
of the important benefits of a PACS program. Although
it is important when building a business case for a PACS
to go beyond purely financial considerations [4], it is also
necessary to work with an organization’s leaders to insure
their financial justifications for a PACS are met.
VENDOR SELECTION
The choice of a PACS vendor is completely unlike the
purchase of an imaging machine. Not only is a PACS
installation probably the biggest purchasing contract a
radiology department will ever undertake, for many vendors, a PACS installation is among their largest and most
complex projects. Complexity is partly the result of rapid
flux in PACS technology; the evolution cycle of technology is very short, and thus product definitions (and consequently the products) change continually. Another difficulty is that the scope of a project can be very difficult to
define. Although vendors define the specific hardware
and software components in their quotes (which are often no more than long lists of such components), this is
only the start of what is needed. Of much more importance is the need to understand the operational functionality of the system. Because each organization is so very
different, it is almost impossible to simply review product
specifications and accurately understand how a product
will work in a particular organization. Site visits to similar
facilities are helpful but cannot provide a true picture of
how a system will work elsewhere. A vendor’s representatives often lack the knowledge or the experience to
understand the daily operations of any particular institution.
Unlike a single system such as a CT scanner, one
cannot “test drive” a PACS. Therefore, in negotiations
with a vendor, it is important to include the following.
●
●
●
●
Detailed specifications and descriptions of functionality of the hardware, the software, all interfaces, and
network requirements. The costs and responsibility for
each interface must be clearly agreed on if unpleasant
surprises and later conflict with the vendor are to be
avoided.
The operational functionality of the system.
Hardware and software upgrade provisions. Vendor
terms such as replacement, upgrade, update, new functionality, and improved functionality have different and
specific definitions, and a purchaser must understand
them. Promises made during presentations must appear in the final contract.
The linkage of hardware with software upgrades. One
must understand that whereas software may be provided free as part of a long-term service contract, ex-
Cohen et al./Planning for PACS 329
●
●
●
pensive hardware upgrades may be needed to run the
new software.
Detailed project documentation, including a project
plan and scope document. The plan should include a
timeline with tasks, task duration, work effort required, and task dependencies. The scope document
will define the goals and objectives of a project by
which success will be measured; it should comprehensively detail the major work areas needed in the project
and issues with which the vendor will assist (and to
what extent). The scope document should define human resources that the vendor will provide and for
how long, including how many days these individuals
will be working at the host site and how many days off
site. One should clearly understand the number and
type of resources the host site must provide and for
how long. A lack of documentation risks management
problems and the potential for unforeseen costs.
Service. It must be clear who will provide the service,
how after-hours coverage will be obtained, and the
delineation between service provided by the vendor
and the host organization.
A detailed educational program. A solid training plan
and quality training materials are as important as the
hardware and software specifications, but the customer
often mistakenly assumes that the vendor will provide
good education. The vendor should provide examples
of its end-user training materials. Because training requirements (e.g., length, timing, and format) may differ greatly between the various user groups, the vendor
training plan should address how each group will obtain the specific competencies necessary to perform its
new duties.
Is a large company or a smaller dedicated PACS provider the best choice? Smaller companies often offer more
flexible functionality at a lower price than large vendors.
However, with a PACS, a long-term vendor relationship
is crucial to ensure continuous system upgrades. Small
companies are more likely either to be unsuccessful and
go out of business or to be purchased by larger vendors
that may discontinue product lines or services. A potentially more expensive but larger vendor may thus be a
better choice.
Purchase options for a PACS are complex and include
leasing, cash payment, shared risk on savings, outsourcing, and paying on a use basis (e.g., per patient or per unit
of storage used). Complex solutions limit the up-front
costs but may cost more in the long term. Because of this
complexity and the high level of buy-in needed by different groups in an organization, PACS purchasing decisions should be delegated to a committee. This committee should include representatives from radiology,
radiology administration, hospital finance, and hospital
IT, as well as a basic PACS physicist. One should involve
potential “power users” of the system, such as orthopedists or neurosurgeons, to ensure that their image management needs will be met as well. An independent
PACS consultant may be needed to assist the committee
if sufficient expertise does not exist in the organization.
TECHNOLOGY ISSUES
Hardware
Because image storage hardware is continually changing,
we do not discuss the various options for storage, other
than to emphasize that a system must provide adequate
storage with rapid access to images. There are many viable options.
However, image viewing monitors provoke considerable debate, with often emotional rather than rational
decision making and thus expensive mistakes. The debate involves choices such as the number of monitors
(two or four), monitor size (19 or 21 inches), and resolution (1K or 2K; ACR standards exist but are controversial). Monitors at 2K resolution provide sharper images
but cost more, produce more heat, are less bright, and are
operationally slower. A four-monitor system is more
costly and loads images slightly more slowly, yet some
radiologists may favor this system because it more closely
resembles the traditional four-panel film viewer. Other
radiologists may prefer the cheaper two-monitor system
because it is faster to electronically change the image
display than to turn one’s head to view four monitors.
Some believe that four-monitor systems are essential for
studies with very large numbers of images, but others
disagree. Yet there are many other important issues besides monitor number and resolution [9], such as the
workstation graphics card and the monitor refresh rate
and brightness. For example, whereas a typical desktop
monitor has a brightness or luminescence of about 100
lumens, high-brightness monitors can have a luminescence of 300 to 600 lumens [9]. Color monitors often
have lower brightness (and resolution) than black-andwhite monitors [9]. Because all users will want the biggest, fastest, and brightest, with all the bells and whistles,
the PACS committee must persuade users that bigger is
not always better; users must define their needs and select
monitors that meet these needs.
In addition to image viewing, each radiologist’s work
area needs additional computer functionality, including
access to the hospital and department information systems and a method of viewing voice dictation reports.
Although these functions can be embedded into the
PACS workstations, it may be easier to install a separate
free-standing computer monitor adjacent to the PACS
monitors.
330 Journal of the American College of Radiology/ Vol. 2 No. 4 April 2005
Software
All PACS vendors provide software allowing many functions to be performed at the workstation, and innovations from one vendor are quickly copied by its competitors. Because functionality is usually more than
adequate, preferences are mostly subjective. It is more
important to consider the ease of the software interface
with hospital information systems, the radiology information system (RIS), and imaging systems, as well as
how the software affects communication between systems. In general, the hospital information system is the
source of patient demographic and billing information,
the RIS sends imaging requests to the PACS and to the
imaging systems, the imaging systems send images to the
PACS, and the PACS tells the RIS that the study is
complete. Signed reports initially enter the RIS, which
forwards the report to the PACS, where the report and
images are integrated. These complex technical details
are not discussed here, but they must be carefully designed and specified. Each interface must be adequately
tested before it is used.
Networks and IT
Radiology equipment purchases are usually controlled
and made by the radiology department, and the equipment usually resides within the department and is maintained exclusively by department personnel. A PACS
project has a much wider scope, requiring significant
involvement of the hospital’s IT staff. Moreover, departmental personnel may lack the technical skills for PACS
administration. Compounding the problem is the need
for the PACS in areas outside of radiology, such as cardiology and endoscopy. For these reasons, IT departments often must take the lead in PACS administration.
Conversely, IT departments may lack the technical information required to manage PACS-RIS integration,
such as DICOM expertise. Therefore, the two groups
must share knowledge and build skill sets related to
PACS technology. Radiology must be willing to yield
significant control to the hospital IT program, whereas
IT personnel must acknowledge the specific needs and
expertise resident within radiology departments. Responsibilities must be clearly defined; in general, radiology should control hardware and software decisions,
equipment location, and workflow issues; IT should control the network and its monitoring, maintenance, and
upgrades, as well as technical support. Both must be
involved in final decisions regarding purchasing, configuration, and implementation.
DISASTER PLAN
The various hospital departments and the vendor should
include a disaster plan as a cost within the initial pur-
chase. The disaster plan has two components: the backup
storage of images in a remote location and a detailed
operational plan if failure of all or part of the PACS
occurs, including the underlying network infrastructure.
The plan should identify equipment to be served by
emergency power supplies. These plans will be complex
and expensive but are crucial.
SECURITY
The technology to provide Health Insurance Portability
and Accountability Act– compliant security within a
PACS (e.g., limiting image access to only those who need
it on a patient-by-patient basis) is available. Many PACS
vendors have not fully implemented it, because implementation is costly and may inhibit workflow. Yet there
is a legal need to limit access to the PACS to maintain
confidentiality, and it should be addressed in the initial
project scope.
SPACE AND ENVIRONMENT ISSUES
The installation of the PACS will eventually free up film
storage space in the radiology department, as film is
phased out. The PACS may free up space in reading
room areas as large alternators are removed and smaller
work stations are installed, but renovation may be required to create optimal reading environments. Yet the
PACS makes huge demands for new space. In every area
where technologists or receptionists work, new space will
be needed for additional computer terminals, as well as
printers, bar code readers, and/or CD burners; this space
may be difficult to find. Special consideration should be
given to the extra space needed during the transition to a
PACS; when historical comparisons are on film and not
the PACS, work areas will be extremely crowded with
new workstations and legacy viewing technology.
Space may be a major issue in inpatient and outpatient
areas, depending on the selection of PACS viewing stations and presence of existing computer equipment. Little new space may be needed in organizations that choose
to rely heavily on Web access to PACS images; this is
especially true if high-end computer workstations are
already in place. However, the installation of new dedicated PACS workstations may be difficult. Decisions on
the type of PACS viewing systems to be used should be
made with consideration of the existing space constraints
in these areas. Space renovation in numerous inpatient
and outpatient areas may add significantly to the overall
PACS project cost.
Whether in the radiology department or on the inpatient wards or clinical outpatient areas, the provision of
low lighting for viewing of images is important. This may
be difficult in the radiology department and almost impossible in very busy clinical areas. In the reading rooms,
Cohen et al./Planning for PACS 331
Table 2. Keys to the success of a picture
archiving and communication system project
implementation
● Strong project manager, dedicated to the
project
● Clear and realistic timetables
● Phased implementation
● Clear definition of individual responsibilities and
accountability
● Comprehensive scope document
● Committed and effective executive
administrative and medical sponsors
● Timely and accountable vendor participation
● Strict adherence to project change
management policies
● Substantial radiologist involvement in the
development of workflows and system design
focused ambient lighting is needed that minimizes reflected light produced by the monitors and other bright
objects [9]. Other environmental considerations in these
areas should include dark wall colors, noise reduction
(through wall, floor, and ceiling treatments), temperature control, and comfortable chairs [10]. Room layout
should allow radiologists to work together yet preserve as
quiet an environment as possible. Full or partial walls
may needed to create individual reading stations. The
actual workspace on the reading station desktop is often
too small [11. It must be adequate for the computer
monitors, keyboard, barcode reader, telephone, and dictation system, and it should have space for paperwork.
Merely removing film alternators and dimming lights in
an existing reading room cannot provide an adequate
PACS work environment.
project is rolled out to a larger audience. The project
manager should be skilled in change management and
have experience in complex project implementation. A
background in industrial process management may be
helpful. The project manager needs some knowledge of
the technology of a PACS but does not need to be an
expert. The keys to implementation include manpower
allocation, project reengineering and planning, communication, and education.
Manpower
Significant manpower resources must be allocated to the
project, from both the host institution and the vendor.
The responsibilities of the host institution and the vendor must be clearly defined to avoid later conflicts and
surprises. The host institution should expect its employees to play a much greater role in the PACS installation
than is usual with other radiology equipment. Resource
needs will vary on the basis of the size of the implementation and the current level of digital technology in the
department. For smaller sites, a single person may fill
multiple roles. Required manpower resources include the
following.
●
PROJECT IMPLEMENTATION
Planning and implementation of a PACS is complex and
should not be underestimated. A large number of tradeoffs must be made. Personnel from many different departments within the hospital must become involved.
Workflow and “cultural” change must be made by many
individuals who are naturally frightened by such change.
The critical success factors of a successful PACS project
are summarized in Table 2. The project plan should
identify all tasks required during the planning, execution,
and closeout phases of the project. Adequate resources
must be allocated to each task and individuals must
clearly understand their own responsibilities. The implementation of a PACS should be phased in so issues and
problems can be identified and resolved before the
●
●
A project manager. Both the host organization and the
vendor must supply a project manager. Ideally, each
should be full-time on the PACS implementation
project. Many health care organizations do not have a
strong pool of project management resources and need
the vendor to assist in this area. Vendors may make
many promises regarding the role that their project
manager will fill because they like to sell project management services. Some vendors deliver a quality
project management service, but others may fail to
deliver the type of project management that most
health care organizations require. One must determine
exactly what aspects of the project the vendor’s project
manager will “own.” Will this person coordinate and
drive all tasks in the project (e.g., communications,
risk, and scope management), or just the ones that the
vendor is directly involved in (e.g., hardware procurement and installation)? The determination of the specific task areas that the host institution will manage
must be made before the actual project planning begins, so that the particular strengths needed can be
assessed.
Network engineers to design and install the required
network infrastructure.
Field engineers to install the PACS equipment and
integrate imaging modalities. Knowledgeable vendorprovided field engineers are crucial to smoothly inte-
332 Journal of the American College of Radiology/ Vol. 2 No. 4 April 2005
●
●
●
●
●
●
●
grating the RIS, imaging systems, and the PACS. Depending on the organization’s level of previous integration, the hospital’s field engineers may not be familiar with all of the required modality settings.
Clinical engineers to research, design, and implement
the integration and upgrades of the imaging systems to
the PACS environment. Clinical engineers should be
directly involved in modality integration. If nonDICOM-compliant modalities are present, they can
be very helpful in identifying the best technology to
use.
“Problem solvers” who have both applications and engineering knowledge.
Radiology department personnel, including administrative directors, technologists, nurses, and clerical
workers. The implementation of a PACS changes the
workflow of almost all departmental activities. Involving key personnel from each functional area early in
the planning process will help ensure a smooth transition.
System administrators to design and build the application, data flow, and databases. They should be identified before the system is built, because it is easier to
learn the system (e.g., database structure, work lists,
filters) if one is involved in the actual system build.
Guessing at system design logic may be time consuming and frustrating. Reliance on a vendor for this service after installation can be costly and may not meet
expectations.
Radiologists to champion the project, make needed
compromises, and contribute to workflow development. They are responsible for selling the project to
administration, referring physicians and other hospital
stakeholders. Limited radiologist input in workflow
development greatly risks the success of the implementation.
Education and communication personnel. Technology-savvy educators are needed to develop training programs, documentation, and tools for a wide variety of
users, including remote users. A marketing person may
be invaluable to communicate the project through a
variety of vehicles to diverse audiences.
A testing team. Because the PACS will likely be integrated with the RIS and imaging systems, the integration relationships must be tested. There must be
project team members who have knowledge of testing
methods and can develop testing scripts, execute the
test transactions, and verify the outcomes. The testing
of all points of system integration and workflow
changes is a critical preparation for production usage.
Process Reengineering and Planning
After project approval, PACS stakeholders will likely
push to move forward quickly. Many times, the pressure
for the immediate gratification of installing hardware and
displaying the images electronically overshadows the
need for true project planning, including process evaluation and reengineering. To be successful, one must understand that the most important phase of the project is
not implementation but planning. Not understanding
this fact makes workflow disruptions, schedule delays,
and budget overruns very likely. Although unexpected
hurdles will be encountered in even a well-planned
project, they are more easier dealt with around a discussion table than after a disruption in production workflow. The PACS technology represents a change in the
way radiology has done business for over 100 years; a
physical element (film) is being taken out of the hands of
physicians, and they are asked to look at a computer
monitor. Anxiety will be high as you ask them to change
their habits and move out of their comfort zone. As
physicians encounter problems in doing their work, they
will question the quality of the project implementation
and the viability of the overall solution. Dedicating adequate resources and time to the planning phase will help
reduce the number of problems encountered by the end
users. Because some issues identified during the planning
phase may not be resolved at system rollout, communicating these outstanding issues to the end-users will let
them know that the problem is known and what is the
planned solution.
The implementation of a PACS requires many processes within radiology and also outside radiology to
be changed. Each component of the value chain of
radiology needs to be carefully analyzed, and the impact of PACS on each component must be understood. These components include study ordering and
scheduling, patient reception, patient flow through
the department, film loan functionality, image generation by technologists, image transfer to radiologists,
radiologists’ workflow, image reporting, and image
and report distribution to referring physicians. In addition, the work patterns of the referring physicians
will change; how they receive images, where they review them, and how they share them with patients
requires careful analysis.
Early in the project, the institution must decide
whether to tackle these problems on its own, to seek help
from the PACS vendor, or to employ skilled consultants.
Consultants can provide a detailed baseline analysis
against which change can be evaluated. Their experience
might also help predict the impact of suggested changes
on outcomes and help optimize the final workflow state.
Consultants also have expertise in other areas, such as
security and benchmarking. Yet consultants add to the
cost of the project, and the quality of their input is extremely variable; many claim to be experts but few are
truly knowledgeable.
Cohen et al./Planning for PACS 333
Although every step of the radiology process needs to
be analyzed, some of the key challenges that will be faced
include the following.
1. When to go filmless. The cheapest and simplest answer is to turn a PACS on and stop the generation of
film on the same day, but this is usually impractical.
The phased implementation of a PACS, starting
within radiology, allows several months of digital
images to accumulate while films continue to be
printed and distributed to referring physicians. This
overlap adds cost but results in a workable archive of
digitized images. Hard copy production can then
end on a certain date, and most subsequently imaged
patients will have had comparison images previously
acquired in digital format. The other solution is a
crash program to quickly digitize all stored films,
which is an unbelievably expensive project that very
few institutions are willing to tackle.
2. Personnel and job duties. Radiology personnel must
understand that their job duties will change. Technologists will rely more heavily on computers to
manage workflow rather than spend time developing film. Film loan personnel may transition to other
duties. New job descriptions must be created so that
staff members understand expectations and accountabilities associated with the new technology
and resultant workflow processes. Achieving the acceptance of new responsibilities can be a challenge,
but buy-in may be stronger if key personnel are
included on the team creating the new job descriptions.
3. A digitization policy for prior and outside films.
Although the digitization of all hard copy storage
films is not needed, each institution should develop
a policy to choose selected images for digitization. A
caveat is that clinicians usually overpredict the number of old films that need digitization. A policy for
digitization of outside films that accompany patients
to the institution should be developed as well.
4. Paper requisitions. If the system is set up such that
paper requisitions are no longer required, workflow
will increase.
5. Voice dictation. If voice dictation does not exist in
the organization, one must decide whether to implement it together with the PACS or to wait until the
PACS is installed and successfully operating. A
PACS makes images very rapidly available to referring clinicians, and they will come to expect reports
to be available just as quickly. However, if voice
dictation is not in use before PACS installation,
because it represents another sharp break in routine,
it is best phased in after the PACS is up and running
successfully.
6. The maintenance of unspecified or broken studies.
The integration of PACS with RIS and the imaging
systems with a DICOM modality work list service
helps ensure the accurate matching of specific studies to specific patients. However, simple errors such
as the misspelling of a patient’s name or entering an
incorrect digit may result in unspecified or “broken”
studies. A staff member must be charged with continually monitoring and resolving these issues.
7. Software usage and workstation operations. Many
decisions must be made about the usage of the PACS
software. Although many functionalities of the
PACS workstation software can be individualized to
specific users, others must be universal. The custom
individualization of software functionality is time
consuming and may require direct collaboration of
application specialists and radiologists. Systemwide
decisions, such as prefetch rules, are difficult and
require consensus and buy-in. These rules decide
which prior studies are fetched, when they are
fetched, and whether they are “pushed” or “pulled”
to the workstation. Prefetch variables include the age
of prior studies, the number of prior studies, “newest
or oldest,” modality, body part, and active studies.
For example, if prior studies will automatically be
displayed with a new study, then “prior studies”
must be defined. If the new study is a chest radiograph, do users want only prior chest radiographs to
be automatically displayed, or do they require all
previous imaging of the chest (e.g., CT, MR)? The
number of studies to be prefetched is also an issue;
increasing the number increases the range of studies
immediately available for comparison but also increases the load placed on the system and may slow
system operation. Another decision that must be
made with prior images is to decide which images to
prefetch and display. For example, one might choose
the six preceding chest radiographs, or the three
preceding chest radiographs and the first three chest
radiographs acquired on that patient, or another of a
seemingly infinite number of combinations. Display
presets are also seemingly infinite and decisions must
be made. What defaults should be used? Should the
defaults be standardized by modality or should there
be a very large number of defaults for each anatomic
area by each modality? What preset windows should
be determined?
8. Work list creation. There are countless ways to display work lists, including by modality, time, patient
location, or image acquisition location. The structure and organization of work lists must meet the
needs of the radiologist. To what degree can they be
individualized? This will vary between different
PACS systems.
334 Journal of the American College of Radiology/ Vol. 2 No. 4 April 2005
Table 3. Picture archiving and communication
system (PACS) implementation, concerns of
referring physicians
Quality of images
● Will image quality be
adequate for my needs?
● Will I get the same
image quality as the
radiologists?
Workflow
● I have a view box in my
consulting room. I like
to show patients images
in my consulting room.
Will I retain this
functionality?
● Who will pull up the
studies?
● Can I print images with
a laser printer?
Workstations
● Will I have the same
functionality as the
radiologist?
● Who will pay for my
workstations?
Availability of prior
● I am used to old films
studies
available in x-ray
jackets. Will prior
studies be available?
System failure
● Electronic system
failures may be more
catastrophic. What
guarantees do I have
that my clinic will not be
shut down by a PACS
failure?
Consultation with
● With PACS I will not
radiologists
spend so much time in
the imaging department.
How will I get good
consultation from my
radiologist?
9. The distribution of viewing stations on inpatient
wards and outpatient clinic areas. Although a simple
computer for Web access takes up little space, is
cheap, and may provide the functionality needed by
most referring clinicians, other clinicians (such as
neurologists) may need the flexibility of dedicated
workstations for viewing complex studies such as
MR. Yet all clinicians will expect to be treated
“equally,” and some of their concerns are summarized in Table 3. There is not a single simple solution.
10. Operating rooms. The provision of a PACS in operating rooms is difficult, and each institution finds
its own solution. Cost and space issues are often
major obstacles to installation. In many institutions,
hard copy film is still provided to surgeons in operating rooms.
11. Clinical conferences. Before the PACS, films were
hand-carried to conference rooms. Inexpensive wallmounted view boxes or overhead projectors allowed
the easy display of images. With the PACS, the ability to display images in multiple conference rooms
scattered throughout the hospital must be envisioned. To encourage clinicians to continue to do
rounds in the radiology department, well-equipped
and comfortable conference rooms should be available there, with wall-mounted image screens to rapidly display images.
12. Clinical histories and other clinical information.
Will radiologists continue to get patients’ histories
from paper requisitions, or through the electronic
system? Will information be displayed on the PACS
workstation or on a separate workstation connected
to the hospital information systems? If the latter,
then additional space must be provided.
Communication
This is an often-neglected component of PACS implementation. The project is complex, time consuming,
lengthy, and involves a huge number of individuals
throughout the organization. It involves cultural and
process changes. It is thus surprising how frequently
communications are neglected. Unfortunately, people
believe that if they just get the hardware installed and the
software up and running, the PACS program will be
successful. This is not necessarily true. It is critical that
referring physicians perceive and believe that all their
needs are being understood and met. Communication
should be a major priority of the PACS implementation
team. A primary communication challenge will be to
help people understand that PACS is not a “radiology”
project but an enterprise-wide project that will positively
affect workflow and patient care beyond the walls of
radiology.
The objectives of a successful communication plan are
to build early excitement and buy-in for the project, set
expectations, and manage information flow to ensure
successful project implementation. To achieve this, there
should be regular communication addressing the following questions:
●
●
●
●
●
What is happening?
Why is it happening?
When is it happening?
How is it happening?
How will it affect me and my patients?
Cohen et al./Planning for PACS 335
●
How will it change the way I work?
Communication should use all available portals, including individual and group meetings, e-mails, attendance at regular hospital organizational meetings, the
development of a PACS informational newsletter, and
the inclusion of PACS items in regular hospital bulletins.
Other portals include Web pages, videotapes, attendance
at grand rounds, and open house demonstrations. Messages need frequent repetition, because individuals may
pay attention only at the time that they feel the message
will directly affect them.
The key to a successful communication program is to
get everyone on board early, anticipate and address concerns proactively, cultivate internal advocates, understand anxieties and needs of constituents, and set and
manage reasonable expectations to ensure acceptance.
End users must feel engaged and must feel that their
specific needs and concerns are being addressed and will
be met. End users’ concerns include the following.
●
●
●
●
●
●
●
What is the PACS?
What is “filmless”?
What will be the cost to me?
Will all my needs be taken care of?
How do I see demonstrations?
What will be the functionality?
How will I get prior images?
One should query end users after several communications to see if they are obtaining needed information and
identify questions or concerns to be addressed in follow-up communications.
Education
The educational process for new radiology equipment
usually involves an applications expert, on site for a few
days, who trains a small number of technologists and
radiologists who usually have a good initial understanding of the technology. Such training is relatively straightforward, inexpensive, and has a limited audience. Education for a PACS is exactly the opposite. Many vendors
have limited experience in PACS training and underestimate its needs and complexity. They thus often follow
educational models that have worked with the sales of
other radiology equipment but are inadequate for a
PACS. Yet education is so critically important in PACS
implementation that it should be a specific item in the
contract negotiated with the vendor, with its own budget
and with clearly defined vendor responsibilities.
Education for a PACS must be viewed as an ongoing
process over many months. Training must be provided at
multiple sites at multiple times, and ongoing education
must be available as new users continually rotate into the
institution. Because poor attendance at training is the
norm, communications must stress the importance of
training. Although training is often provided before final
implementation, the most valuable training occurs after a
PACS is operational. It is only then that the user may
become engaged and motivated to learn.
It is critical that training support materials be provided, which should include responses to frequently
asked questions. Flip charts or other simple guide material should be made available at every workstation, and a
mechanism provided to ensure that this material is removed from the workstation. A well-staffed help hotline
must be available. More detailed educational material
must be available to end users who want information on
customizing the functionality of the system.
Trainer selection is difficult. Should one provide individuals with superficial but very broad-based knowledge
of the PACS, or with more focused, very detailed knowledge so that the needs of the most demanding users can
be met? The answer is probably a compromise, with a
need to produce both types of training in different circumstances. Training is often needed in multiple locations, and an effective way to train a large number of
users is to provide tiers of training. There can be demonstration classrooms set up for large numbers of new users
or those wanting to view the system for the first time.
User-specific training can take place at the site of actual
use of the system. If training resources are at a premium,
general training can be accomplished by training CDs,
videos, or well-written and self-directed training manuals. A list of all individuals to be educated should be
generated, and careful tracking should ensure that all of
the end users are in fact trained. Although some vendors
desire to develop expert trainers within the organization,
this should not substitute for the other components of a
good educational program that have been described
above.
SUCCESS CRITERIA
A project can be measured in many ways: by procedures
or timelines met, by the deliverables achieved, or even by
the project team’s perception. It is important that the
project plan initially defines not only the project scope
and goals but also how the success of the project will be
measured. Success criteria should be specific, measurable,
and time based. It is very important to celebrate as specific objectives are completed.
MISTAKES, PITFALLS, AND PROBLEMS
The following is a list of potential hazards. Awareness of
these traps may help a new PACS purchaser avoid them.
336 Journal of the American College of Radiology/ Vol. 2 No. 4 April 2005
Vendor Issues
●
●
●
●
●
●
●
●
●
●
Believing the vendors and not having items specifically
addressed in the signed purchase contract. Meaningful
incentives and penalties are key to motivated vendor
participation.
Inadequate specificity in the contract.
The absence of a well-written, comprehensive scope
document.
The failure to uphold initial project scope (scope
creep). The project may continue to grow beyond its
initial definitions and budget.
The failure to understand that vendors are very good at
selling traditional x-ray equipment and even hardware
and software but may have limited experience in process change and management.
Not specifying a clear plan for upgrades, software enhancements, and future product releases in the contract.
The failure to appreciate that no vendor product is
perfect and that all vendors are continually trying to
improve their products.
Not understanding that the vendor cannot customize
improvements to meet each individual customer’s needs.
Overestimating the return on the investment of the
project and the time frame in which in will be
achieved.
Believing that software will function as it is marketed.
Technology Issues
●
Focusing too much on the technology and not enough
on the people.
Space and Environment Issues
●
Insufficiently evaluating the type, location, and distribution of viewing stations needed in the hospital
wards, hospital outpatient clinics, and private practitioner offices.
●
●
●
●
●
●
●
●
Believing that best practice solutions will be adopted
by your organization. Frequently reinventing the
wheel, when best practice solutions may be available
from other organizations.
Not insisting on initial default operational settings
when the vendor provides too much flexibility.
Underestimating the number and complexity of problems arising after PACS use begins.
Not implementing a mechanism of ongoing system
audits and preventive maintenance programs.
A lack of understanding of the complexity of the development of body part use and work lists.
Not insisting on greater radiologist involvement in the
workflow design.
Underestimating the level of the computer literacy of
end users.
Not creating a hospital-wide viewing solution standard; allowing referring physicians to choose their
technology.
Communication Issues
●
●
●
●
●
●
●
●
Not understanding that all of the physician end users
desire top-of-the-line monitors, particularly if they are
not paying for them.
Allowing end users to develop unrealistic expectations
beyond those that can be provided and beyond those
that already exist in their practices.
The failure to provide an adequate educational program and an excellent help desk.
Not understanding that a single product cannot please
everyone.
The compromise of training schedules by trainee noshows.
Underestimating the anxiety of potential end users.
Underestimating original radiologist resistance to
change.
Allowing too many people to influence project
direction.
Implementation Issues
●
●
●
●
●
Recognizing that because an enterprise-wide PACS
program from initial concept to final implementation
may take several years, there may be significant turnover of involved individuals both in the host organization and in the vendor.
Underestimating the complexity of integration of the
RIS with the PACS, imaging modalities with the
PACS, and imaging modalities with the RIS.
Underestimating hospital manpower needs.
Underestimating the complexity of finding a good solution for the operating rooms.
The lack of process benchmarks and data and the lack
of industry best practices. Do not rely on the vendor
for these.
CONCLUSION
Every facility is different,and every PACS project will be
different. The current state of technology, institution
size, skill sets in place, project goals, and financial resources all affect a new technology implementation
project such as a PACS. The aim of this article is to
identify many of the critical factors for success in planning for a PACS. However, the implementation of a
successful PACS project is a giant undertaking; not every
situation or issue can be anticipated. Through careful
planning, coupled with attention to all the issues outlined in this review, a PACS project can become fun and
hugely rewarding.
Cohen et al./Planning for PACS 337
REFERENCES
1. Samei E, Seibert JA, Andriole K, et al. AAPM/RSNA tutorial on equipment selection: PACS equipment overview: general guidelines for purchasing and acceptance testing of PACS equipment. Radiographics 2004;
24:313-34.
6. Don S, Albertina MJ, Ammann D. Soft-copy sonography: cost reduction
sensitivity analysis in a pediatric hospital. Am J Roentgenol 1998;170:
571-5.
7. Reiner BI, Siegel EL, Hooper FJ, Glasser D. Effect of film-based versus
filmless operation on the productivity of CT technologists. Radiology
1998;207:481-5.
2. Reiner B, Siegel E. Understanding financing options for PACS implementation. Picture archiving and communication system. J Digit Imaging
2000;13:49-54.
8. Langlotz CP, Seshadri S. Technology assessment methods for radiology
systems. Radiol Clin North Am 1996;34:667-79.
3. Channin DS, Chang PJ. Primer on computers and information technology, part five: glossary. Radiographics 1997;17:1310-5.
9. Wang J. Human perception of digital displays. Diagn Imaging (San
Franc) 2000;(suppl):13-9.
4. Diamond D, Wasilewski D. Purchasing strategies: go beyond financials in
making case for PACS. Diagn Imaging (San Franc) 2000;(suppl):27-32.
10. Ratib O, Valentino DJ, McCoy MJ, Balbona JA, Amato CL, Boots K.
Computer-aided design and modeling of workstations and radiology
reading rooms for the new millennium. Radiographics 2000;20:1807-16.
5. Pratt HM, Langlotz CP, Feingold ER, Schwartz JS, Kundel HL. Incremental cost of department-wide implementation of a picture archiving
and communication system and computed radiography. Radiology 1998;
206:245-52.
11. Pilling JR. Lessons learned from a whole hospital PACS installation.
Picture archiving and communication system. Clin Radiol
2002;57:784-8.